Clustering by K-Means Method and K-Medoids Method

Author:

Önay Onur1

Affiliation:

1. School of Business, Istanbul University, Turkey

Abstract

Data science and data analytics are becoming increasingly important. It is widely used in scientific and real-life applications. These methods enable us to analyze, understand, and interpret the data in every field. In this study, k-means and k-medoids clustering methods are applied to cluster the Statistical Regions of Turkey in Level 2. Clustering analyses are done for 2017 and 2018 years. The datasets consist of “Distribution of expenditure groups according to Household Budget Survey” 2017 and 2018 values, “Gini coefficient by equivalised household disposable income” 2017 and 2018 values, and some features of “Regional Purchasing Power Parities for the main groups of consumption expenditures” 2017 values. Elbow method and average silhouette method are applied for the determining the number of the clusters at the beginning. Results are given and interpreted at the conclusion.

Publisher

IGI Global

Reference72 articles.

1. Ahi, L. (2015). Veri Madenciliği Yöntemleri İle Ana Harcama Gruplarının Paylarının Tahmini [Estimation Of Main Expenditure Groups’ Portion With Data Mining Methods] (Unpublished master’s thesis). Hacettepe Üniversitesi Fen Bilimleri Enstitüsü.

2. Veri Madenciliği Yöntemleri ile Ülkeleri Gelişmişlik Ölçütlerine Göre Kümeleme Üzerine Bir Uygulama [An Application On Clustering Countries With Data Mining Methods Based On Development Criteria].;B.Akkuş;AURUM Mühendislik Sistemleri ve Mimarlık Dergisi,2019

3. Bölgesel kalkınma politikalarının etkinliği: Türkiye alt bölgeler bazında bir uygulama [The Effectiveness of Regional Development Policies: An Application on Sub-Regions of Turkey].;F.Bakırcı;Atatürk Üniversitesi Sosyal Bilimler Enstitüsü Dergisi,2014

4. Clustering with Bregman divergences.;A.Banerjee;Journal of Machine Learning Research,2005

5. Accelerating k-medoid-based algorithms through metric access methods

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3